Kurzfassung
Nicht verfügbar

Abstract
An approach to automated feature extraction is presented which uses an object-oriented geodata model as the framework to store contextual knowledge and to use this both to control feature extraction routines and to automatically differentiate between linear feature classes (roads, railways, rivers etc.). The problem of geographic extraction has proved complex and ideally requires the incorporation of contextual clues similar to those used by human interpreters of imagery. The paper describes a proof of principle system developed under UK Ministry of Defence Corporate Research funding. The geodata model comprises a class hierarchy representing the features under study and their likely relationships. Each class of object within this model contains criteria that need to be satisfied in order to strengthen the belief that an instance of that object type has been recognised. The system described has at its core a spatially enabled object oriented database. This enables the extraction of linears to be divorced from the classification process which gives the system the flexibility to build up evidence of class membership from a variety of sources. In this way linears can be tagged with initial probabilities of class membership and refined following further processing, such as network building stages, where classification conflicts are identified and resolved to provide more probable class memberships.